中文摘要：

针对平衡数字水印技术中的鲁棒性和不可感知性进而解决图像的版权保护问题,提出了一种基于Contourlet变换和QR分解的鲁棒性盲水印算法。选用有意义的二值图像作为待嵌入的水印,首先对原始图像进行Contourlet变换得到低频子带,接着对低频子带进行分块,然后对得到的子块进行QR分解;最后二值水印采用量化索引调制（QIM）方法重复嵌入到R矩阵的第1行中。实验结果表明,本文算法能有效抵抗常规信号处理操作的攻击,如对裁剪、高斯噪声、椒盐噪声、乘性噪声、均值滤波、高斯滤波和JPEG/JPEG 2000压缩攻击具有很强的鲁棒性,并且用到的为QR分解,所以在计算复杂度方面远低于基于SVD分解的水印算法。水印的提取过程中没有借助原始图像的任何信息,属于盲水印算法。因此,本文提出的数字水印方法可在数字产品版权保护和多媒体信息传输方面推广。

英文摘要：

To achieve a better trade-off between robustness and imperceptibility and solve the problem of copyright protection,A robust blind digital watermarking scheme based on QRdecomposition and contourlet transform is developed in this paper.The meaningful binary image is selected as the watermark.Firstly,the original image is decomposed by contourlet transform,and the obtained Contourlet coefficients are divided into sub-blocks.Then,QRdecomposition is applied to the sub-blocks.Finally,the watermark is embedded into the first column of Rmatrix with quantization index modulation（QIM）method.Experimental results show that the proposed algorithm is robust enough to resist the common signal processing operations,such as cutting,Gaussian noise,salt and pepper noise,multiplicative noise,mean filtering,Gaussian filtering and JPEG/JPEG 2000 compression,and the approach has a lower calculation complexity compared with singular value decompostion（SVD）based methods.What′s more,the proposed watermarking method can realize the blind extraction of watermark without any original image message.Therefore,the scheme proposed in this paper can be widely accepted in the field of digital products copyright protection and multimedia information transmission.